The present technology relates to a power management device, a power management method, and a demand notifying device.
In the past, power supply to power consumers in many countries has been monopolized by power companies. However, the entries of entities other than power companies into a power retailing business and power futures trading have recently been realized or promoted with an objective of further improving efficiency of the electric power industry and supplying power at lower prices by introducing principles of competition into the electric power industry.
Power futures trading for example predicts a necessary amount of power in advance and sells or buys power for a following day or up to 24 hours ahead through a power market. Thus, in order for entities engaged in power retailing and power futures trading to have an advantage in power trading and make large profits, power demand needs to be predicted accurately.
As technology for predicting power demand, a total power demand amount predicting device has been proposed which predicts a total amount of power demand by taking in meteorological variables including an air temperature and humidity in the past and data on a total amount of power demand and further performing learning by a neural network (Japanese Patent Laid-Open No. Hei 5-18995 referred to as Patent Document 1 hereinafter).